A New Approach Against Sugar Cataract Through Aldose Reductase Inhibitors

Exp. Eye Res. (1999) 69, 533–538 Article No. exer.1999.0729, available online at http:\\www.idealibrary.com on

A New Approach Against Sugar Cataract Through Aldose Reductase Inhibitors S T E F A N I A B A N D I T E L L Ia, E N R I C O B O L D R I N Ib, P I E R G I U S E P P E V I L A R D Oa, I L A R I A C E C C O N Ia, M A R I O C A P P I E L L Oa, M A S S I M O D A L M O N T Ea, I S A B E L L A M A R I N Ia, A N T O N E L L A D E L C O R S Oa    U M B E R T O M U R Aa* UniversitaZ di Pisa, Dipartimento di Fisiologia e Biochimica, Laboratorio di Biochimica, via S. Maria, 55, 56100, Pisa, Italy and b Farmigea S.p.A., Research Division, via Carmignani 2, 56127 Pisa, Italy

a

(Received Rochester 20 April 1999 and accepted in revised form 30 June 1999) Aldose reductase inhibition is one of the therapeutic strategies that has been proposed to prevent or ameliorate long term diabetic complications including retinopathy and sugar cataract. Rats were fed with a galactose rich diet and the aldose reductase inhibitor Tolrestat was topically delivered by ocular instillation. The levels of lens aldose reductase activity, galactitol and the onset of cataract were evaluated during and after treatment with the inhibitor. Topical application of 1–3 % Tolrestat (10 µl) four times daily resulted, after 9 days, in a significant decrease in the enzyme activity. Well after interrupting treatment with the drug, the enzyme activity remained impaired and galactose induced cataract was prevented. Our findings may represent the basis for therapeutic plans to prevent sugar cataract by long term cyclic treatments with aldose reductase inhibitors, with reduction in drug doses and side effects. # 1999 Academic Press Key words : Aldose reductase inhibitors ; diabetic cataract ; drug delivery ; lens ; tolrestat.

1. Introduction The hyperosmotic theory of the etiology of sugar cataract identifies aldose reductase (ALR2, EC 1.1.1.21) as the primary factor responsible for this pathological alteration induced by hyperglycemic or hypergalactosemic conditions. Increased formation of both sorbitol and galactitol by the NADPH-dependent reduction of glucose and galactose, respectively, would seem to cause an intramolecular osmotic imbalance which promotes the occurrence of a cascade of events leading to lens opacification (Kinoshita, 1974). Alternative hypotheses for the genesis of sugar cataract have also been proposed, such as those citing the oxidative stress induced by hyperglycemic conditions (Jones and Hotehersall, 1993) and\or protein glycation phenomena (Stevens et al., 1978 ; Baynes, 1991). These data would seem to show that cataract aetiology and development is a multifactorial event. However, ALR2 has become the target of a great deal of research into finding specific and efficient inhibitors of the enzyme. Even though there is no clear evidence of drugs preventing or blocking complications of diabetes in humans, the effectiveness of aldose reductase inhibitors (ARIs) was successfully tested in various animal model systems as potential agents to prevent sugar cataracts as well as other pathological disorders connected to hyperglycemic conditions

* Address correspondence to : Umberto Mura, Dipartimento di Fisiologia e Biochimica, Laboratorio di Biochimica, via Santa Maria, 55-56100 Pisa, Italy.

0014–4835\99\110533j06 $30.00\0

(Sarges and Oates, 1993 ; Tomlinson et al., 1992). Moreover, mice, which are essentially devoid of lens ALR2, do not develop sugar cataract in hyperglycemic conditions or when subjected to a galactose rich diet (Varma and Kinoshita, 1974). Transgenic mice, which over express ALR2 do develop cataract in hyperglicemic conditions (Lee et al., 1995). Thus, it appears that the prevention of polyol accumulation by inhibiting ALR2 is correlated to the prevention of cataracts and other ocular diabetes–associated pathologies (Beyers-Mears et al., 1996 ; Sato et al., 1998 ; Robison et al., 1989). However, antioxidants, including some ARIs, were suggested to be a relevant factor in preventing the onset of cataract (Jiang et al., 1991 ; Linklater et al., 1986 ; Wolff and Crabbe, 1985). Although an osmoregulatory role has been proposed for ALR2 in the kidney, the specific function of this enzyme is still an open issue. Recent evidence seems to show that the enzyme has a detoxification role against toxic aldehydes derived from lipid peroxidation (Srivastava et al., 1995 ; Vander Jagt et al., 1995). If so, a general impairment of ALR2 activity (such as the one predictable by systemic treatment with ARIs) may clearly not always be beneficial. It may therefore be necessary to evaluate the balance between the advantages derived from the control of the level of the polyols and a reduced efficiency in antagonizing oxidative damage. Proposed as a preventive therapy, ALR2 inhibition is easily predictable as a long term treatment and special care in the evaluation of possible side effects is required. Indeed, there are reports (Sorbinil Retinopathy Trial Research Group, 1990 ; Krans, 1992) of the occurrence of side # 1999 Academic Press

534

effects during ARIs treatments which impaired, in some instances, the accomplishment of clinical trials. The results reported here on the topical treatment of galactose fed rats with Tolrestat, a well known ARI, reveal the potential of an effective enzyme inhibition with a significant reduction in drug doses. 2. Materials and Methods Materials NADPH, NADH, D,L-glyceraldehyde, GSH, dithiothreitol (DTT) and milk xanthine oxidase (EC 1.1.3.22) were purchased from Sigma Chemical Co. (St. Louis, MO, U.S.A.). YM-10 ultrafiltration membranes were from Amicon, Inc. and Sephadex G-75 from Pharmacia Biotech. Inc. (Uppsala, Sweden). N-o[5-(trifluoromethyl)-6-methoxy-1-naphthalenyl] thioxomethylq-N-methylglycine (Tolrestat) from Ayerst Research Laboratories (Princeton, NJ, U.S.A.) was solubilized at proper final concentrations in a delivery medium (Farmigea SpA, International Patent Pending No PCT\IT98\00203), pH 7n4. All other chemicals were of reagent grade. Calf lenses for ALR2 purification were obtained from freshly slaughtered animals at a local slaughterhouse (IN.AL.CA. S.p.A., Castelvetro, Modena). Three week old male Sprague-Dawley rats, housed in accordance with the National Institutes of Health guidelines, were kept under a normal diet (TRM from Harlan Tekland, U.K.), and treated according to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. In order to induce sugar cataract, the above diet was supplemented with 50 % galactose.

S. B A N D I T E L L I E T A L.

barbital and the eyes enucleated. Lenses were removed by dissection using a posterior approach and the vitreous, iris, and ciliary bodies were removed. Lenses were quickly washed in saline solution, weighted, and individually homogenized at 4mC in approximately eight volumes of 10 m sodium phosphate buffer pH 7. The suspension was centrifuged for 30 sec in a Beckman Microfuge E and the supernatant (crude extract) was used to measure enzyme activities and protein content. Other Methods Galactitol was measured by gas chromatography on lens samples prepared according to Manius et al. (1972). Aldose reductase was purified from rat and bovine lenses as previously described (Del Corso et al., 1990). Protein concentration was estimated by the Coomassie blue binding assay (Bradford, 1976) with bovine serum albumin as standard. The incidence of cataract (matured nuclear cataract) (Ao et al., 1991) in the animals subjected to the galactose rich diet was evaluated by slit-lamp examination. 3. Results Aldose reductase activity, measured in crude lens extracts after eye topic treatment of rats with Tolrestat, progressively declined to an extent proportional to the ARI concentration (Fig. 1). After 9 days of treatment with 1, 2 and 3 % of the ARI, a residual activity of 74, 40 and 17 %, respectively, was measured. After

Assay of Enzyme Activities Aldose reductase activity was spectrophotometrically measured at 37mC as previously described (Cappiello et al., 1994) by using 4n7 m D,Lglyceraldehyde as substrate. Sorbitol dehydrogenase (EC 1.1.1.14) activity was spectrophotometrically measured at 37mC, as previously described (Marini et al., 1997), by using 0n4  D-fructose as substrate. Purine nucleoside phosphorylase (EC 2.4.2.1) activity was spectrophotometrically measured at 37mC, as previously described (Barsacchi et al., 1992), by using 0n5 m inosine as substrate and milk xanthine oxidase as ancillary enzyme. Glutathione S-transferase (EC 2.5.1.18) activity was measured at 25mC, as previously described (Dal Monte et al., 1998), by using 1 m of both GSH and 1-chloro-2,4-dinitrobenzene as substrates. In all cases, one unit of enzyme activity is the amount of enzyme that catalyses the formation of 1 µmol of product min−" in the assay conditions indicated. Preparation of Lens Extract Rats were killed by an overdose of sodium pento-

F. 1. Inhibitory effectiveness of Tolrestat topic treatment on rat lens aldose reductase. Rats were subjected to topical applications of different Tolrestat solutions by instillation of a 10 µl drop per eye four times a day for 9 days (arrow a). The treatment was then suspended for 18 days and then resumed for 4 more days (arrow b). The symbols (#), ($), ( ) and (>) refer to 0, 1, 2 and 3 % of Tolrestat solutions, respectively. Aldose reductase activity (mU mg−" of protein) was measured as described in the Materials and Methods Section.

CATARACT AND ALDOSE REDUCTASE INHIBITION

535

weeks after the treatment. Moreover, the recovered activity was susceptible to Tolrestat when the drug treatment was resumed (Fig. 1, arrow b). Lens enzymes, such as sorbitol dehydrogenase, purine nucleoside phosphorylase and glutathione S-transferase, chosen as control enzyme activities, were essentially unaffected by the ARI-treatment (data not shown). The prolonged impairment of ALR2 activity was also observed when rats received the drug while following a galactose rich diet for 90 days (Fig. 2). In this case after 18 days of Tolrestat treatment, aldose reductase remained in a low activity state for quite a long time. While a recovery of 50 % of the initial enzyme activity was observed approximately 2 months after the 2 % Tolrestat treatment ceased, only 15 % of the initial ALR2 activity was observed after the same period of time in the lenses of animals subjected to 3 % drug treatment. The reduction in the galactitol level in the lenses of animals on a galactose rich diet, which were also receiving the ARI, is indicative of the effectiveness of the enzyme inhibition in vivo (Table I). While 1 % Tolrestat did not appear to significantly affect the galactitol formation, 2 and 3 % Tolrestat treatment markedly reduced, though did not elim-

F. 2. Inhibitory effectiveness of Tolrestat topic treatment on lens aldose reductase in rats subjected to galactose rich diet. Rats, as described in Fig. 1, were subjected to 18 days of Tolrestat treatment while being fed with a 50 % galactose containing diet. At the indicated times lenses were removed and ALR2 activity (mU mg−" of protein) measured on crude extracts. The symbols (#), ($), ( ) and (>) refer to 0, 1, 2, 3 and 3 % of Tolrestat solutions, respectively.

suspension of the drug treatment, the enzyme activity recovered very slowly. Following the treatment with 3 % Tolrestat solution, the recovery of the enzyme activity did not reach 50 % of the initial value until 2

T I Level of galactitol ( µmol g−" lens wet weight) in the lens of galactose fed animals subjected to topic Tolrestat treatment Days on diet

Galactose

Galactosej Tolrestat 1 %

Galactosej Tolrestat 2 %

Galactosej Tolrestat 3 %

0 4 8 15 17

0n00p0n0 (6) 1n41p0n21 (6) 1n60p0n22 (6) 1n79p0n17 (6) 0n32p0n31 (8)

0n00p0n0 (6) 1n50p0n12 (5) 1n74p0n18 (5) 1n36p0n37 (3) 0n135p0n05 (2)

0n00p0n0 (6) N.D. 1n24p0n18 (4) 0n61p0n11 (2) N.D.

0n00p0n0 (8) 1n15p0n37 (8) 0n78p0n25 (10) 0n64p0n07 (4) 0n88p0n17 (4)

N.D. : Not determined. Values represent the meanp.. of independent measurements. The number of treated animals is reported in brackets.

T II Effect of Tolrestat topic treatment on the cataract development in galactose fed rats. Rats as reported in Fig. 1, fed with a galactose rich diet, were subjected to four applications per day of 10 µl drops of 0 (control), 1, 2, and 3 % Tolrestat and the occurrence of cataract was evaluated by slit lamp examination Cataract occurrence (%)* Days on diet 12 13 14 15 16 17 18 19 20 21

Galactose 8 8 25 52 61 72 90 93 97 100

(13) (13) (36) (33) (33) (29) (29) (29) (29) (18)

Galactosej Tolrestat 1 %

14 32 32 33 44 44 50 50

(21) (19) (19) (18) (18) (18) (18) (18)

Galactosej Tolrestat 2 %

0 0 0 0 0 0 0 0

GalactosejTolrestat 3%

(8) (7) (7) (7) (7) (7) (7) (7)

* Values are cumulatives of three experiments ; the number of treated animals is reported in brackets.

0 0 0 0 0 0 0 0

(33) (31) (31) (29) (29) (29) (29) (18)

536

S. B A N D I T E L L I E T A L.

F. 3. Flow chart of the onset of sugar cataracts in galactose fed rats undergoing Tolrestat treatment. Rats, as described in Fig. 1, fed with a 50 % galactose containing diet ( ) were subjected to either 1 ( ) or 3 % ( ) Tolrestat topic treatment and cataracts were evaluated by slit lamp examination after 25 days. Non cataractous animals for each Tolrestat concentration were split into two groups which were fed, in the absence of the drug, either with a galactose rich diet or with a normal diet ( ). The figures in brackets (xc\xt) at different times, refer to the number of cataractous animals (xc) over the total number of treated animals (xt).

inate, the polyol formation. In fact, the galactitol level was not reduced below 50 % of the proper control value of animals which only received the delivery medium with no drug. The control value on day 17 of the diet is not included in this evaluation. In fact at that time, the level of lens galactitol in both control and 1 % Tolrestat treated animals decreased abruptly, probably as a result of leakage from cataractous lenses. When rats were analysed for the onset of cataract (Table II), it was clearly evident that the treatment with 2 and 3 % Tolrestat was very effective against cataract development. Cataract did not develop in these conditions after 3 weeks of treatment. The ocular application with 1 % of the drug delayed and reduced, but did not eliminate, the onset of cataract. Fig. 3 shows the incidence of cataract occurring in rats

after ceasing the drug treatment. The results related to the 3 weeks of Tolrestat treatment concur with those shown in Table II. In particular, 100, 50 and 0 % of the animals developed cataract after treatment with 0, 1 and 3 % of Tolrestat solutions, respectively. After suspension of the drug treatment, animals that did not have cataract were split into two groups which were housed by receiving either normal or galactose rich diets. While within a week all the animals previously treated with 1 % Tolrestat developed cataract, no cataract was observed for at least 45 days among rats previously treated with 3 % Tolrestat, irrespective of their diet. The ability of ALR2 to retain bound Tolrestat was tested making use of the purified enzymes isolated from both rat and bovine lenses. The purified enzymes

CATARACT AND ALDOSE REDUCTASE INHIBITION

537

were incubated (7–15 µ) at 25mC in the presence of a stoichiometric amount of Tolrestat for 30 min, which led to a complete loss of ALR2 activity. The inactive enzyme preparations were either extensively dialysed on Amicon YM10 membrane against 10 m sodium phosphate buffer pH 7n0 containing 2 m DTT, or subjected in the same buffer to gel filtration chromatography on Sephadex G-75. No recovery of enzyme activity was observed. When the purified enzymes were incubated in the above conditions for 30 min in the absence of the ARI, the recovery of enzyme activity ranged from 70–90 %.

Tolrestat developed cataract, none of the animals receiving the 2 or 3 % Tolrestat topic applications developed cataract during the treatment. In fact, no cataracts were observed in the galactose-Tolrestat treated animals, even when the treatment was interrupted while galactose diet was continued (Fig. 3). These results support the hypothesis that ALR2 is involved in the onset of cataracts (Fig. 2). Nevertheless, it is worth noting that a low, but still well detectable, galactitol level is present in those animals in which no cataract developed (2 and 3 % Tolrestat treated animals). This line of evidence, which is supported by previous results (Dvornik et al, 1973 ; Unakar et al., 1992 ; Datiles et al., 1982 ; Beyers-Mears et al., 1985) keeps open the debate on the unity of the osmotic imbalance in the aetiology of sugar cataract (Crabbe and Goode, 1998 ; Jiang et al., 1991 ; Linklater et al., 1986 ; Wolff and Crabbe, 1985) and indicates that a galactose level as high as approximately 0n7 µmol g−" of tissue, is in any event well tolerated in rats lens, and is not sufficient to induce cataract. Whether the high prolonged inactivation status of ALR2 derives from an apparent long bioavailability of the drug in the lenticular tissue, or from the inhibitory efficiency of the drug due to tighter binding to the enzyme, remains to be defined. However, the tight binding ability of Tolrestat to the enzyme must be a relevant factor in extending the ALR2-depressed state of the lens. In fact, both extensive dialysis and efficient gel filtration chromatographies were unable to rescue the enzyme activity from purified rat and bovine lens ALR2, once treated with Tolrestat. It is clear that these results are due to the availability of a highly selective and potent tight binding inhibitor and, referred to the in vivo results, to an efficient drug delivery system, able to target the enzyme locally. The prolonged impairment of the enzyme activity as reported in this study may well allow cycles of drug topic treatment to be designed with a significant reduction in the drug doses and number of eye applications. Consequently, it might be feasible to plan clinical trials of sufficient time length in order to assess the potential advantages of ARIs treatment (Raskin, 1992). Such an approach may reduce the risk of side effects without any loss of efficiency in ALR2 inhibition, thus opening new strategies for sugar cataract prevention.

4. Discussion The usefulness of ARIs in the prevention or amelioration of human diabetic complications is still an open issue. The need for long term treatments is a serious impairment for a successful accomplishment of clinical trials. Any new device or strategy that could improve the inhibitory action of ARI would naturally be a key to success. In this regard the control of drug delivery to the target organ is clearly as relevant as the effectiveness of the drug inhibiting the target enzyme. Eye topical treatment is certainly a less intrusive and widespread approach to solve the problem of sugar cataract than a systemic treatment. However, it can be therapeutically effective only if the drug is made truly permeable in the eye lens and thus easily available for the target enzyme. The low solubility and bioavailability of Tolrestat, a well known ALR2 inhibitor which was used in this work, was overcome by the use of an optimized solubilizing cocktail which allows an easy and efficient delivery of the ARI to the lenticular tissue. Rats were thus subjected to an eye topic treatment by receiving droplets of Tolrestat solutions. In these conditions, ALR2 was readily and specifically inhibited (Fig. 1), while other unrelated enzymes, namely sorbitol dehydrogenase, purine nucleoside phosphorylase and glutathione S-transferase, used as reference controls, were essentially unaffected by the treatment (data not shown). As expected, both the rate and the extent of ALR2 inhibition appeared to be dependent on the inhibitor concentration. However, what was new in our results was the long impairment of ALR2 after the treatment had been stopped. Moreover, once it had been recovered after Tolrestat treatment, ALR2 was against susceptible to the inhibitory action of the drug (Fig. 1, arrow b). An even higher effectiveness of Tolrestat in inhibiting ALR2 was observed when animals received the drug for 18 days, while being fed with a galactose enriched diet (Fig. 2). Again, after the ARI treatment was stopped, ALR2 appeared to be inhibited for quite a long time. In this case, the impairment of the enzyme activity and thus the decrease in lens galactitol levels (Table I), was associated with a significant reduction in cataract formation. Indeed, as reported in Table II, while within 3 weeks all galactose fed rats which did not receive

Acknowledgements This work was supported in part by a Grant from Pisa University and from the Italian Board of Education (MURST) and in part by Farmigea S.p.A, Pisa, Italy.

References Ao, S., KiKuchi, C., Ono, T. and Natsu, Y. (1991). Effect of instillation of aldose reductase inhibitor FR74366 on Diabetic Cataract. Invest. Ophthalmol. Vis. Sci. 32, 3078–83.

538

Barsacchi, D., Cappiello, M., Tozzi, M. G. et al. (1992). Purine nucleoside phosphorylase from bovine lens : purification and properties. Biochim. Biophys. Acta. 1160, 163–70. Baynes, J. W. (1991). Role of oxidative stress in the development of complications in diabetes. Diabetes 40, 405–12. Beyers-Mears, A., Kelly, K. and Cruz, E. (1985). Synergism of Sorbinil and normal diet on reversal of stage-II sugar cataract. Pharmacology 31, 170–9. Beyers-Mears, A., Mistry, K., Diecke, F. P. J. and Cruz, E. (1996). Zopolrestat prevention of proteinuria, albuminuria, and cataractogenesis in diabetes mellitus. Pharmacology 52, 292–302. Bradford, M. M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizating the principle of protein-dye binding. Anal. Biochem. 72, 248–54. Cappiello, M., Voltarelli, M., Giannessi, M. et al. (1994). Glutathione dependent modification of bovine lens aldose reductase. Exp. Eye Res. 58, 491–501. Crabbe, M. J. C. and Goode, D. (1998). Aldose reductase : a window to the treatment of diabetic complications ? Prog. Retinal Eye Res. 17, 313–83. Dal Monte, M., Cecconi, I., Buono, F., Vilardo, P. G., Del Corso, A. and Mura, U. (1998). Thioltransferase activity of bovine lens glutathione S-transferase. Biochem. J. 334, 57–62. Datiles, M., Fukui, H., Kuwabara, T. and Kinoshita, J. H. (1982). Galactose cataract prevention with Sorbinil, an aldose reductase inhibitor : a light microscopic study. Invest. Ophthalmol. Vis. Sci. 22, 174–9. Del Corso, A., Barsacchi, D., Giannessi, M., et al. (1990). Bovine lens aldose reductase : tight binding of the pyridine coenzyme. Arch. Biochem. Biophys. 283, 512–8. Dvornik, D., Simard-Duquesne, N., Krami, M., et al. (1973). Polyol accumulation in galactosemic and diabetic rats : control by an aldose reductase inhibitor. Science 182, 1146–8. Jiang, Z. Y., Zhou, Q. L., Eaton, J. W., Koppenol, W. H., Hunt, J. V. and Wolff, S. P. (1991). Spirohydantoin inhibitors of aldose reductase inhibit iron- and coppercatalysed ascorbate oxidation in vitro. Biochem. Pharmacol. 42, 1273–8. Jones, R. H. V. and Hotehersall, J. S. (1993). The effect of diabetes and dietary ascorbate supplementation on the oxidative modification of rat lens βL-crystallins. Biochem. Med. Metabol. Biol. 50, 197–209. Kinoshita, J. H. (1974). Mechanism initiating cataract formation. Invest. Ophthalmol. Vis. Sci. 13, 713–24. Krans, H. M. J. (1992). Recent clinical experience with aldose reductase inhibitors. J. Diabetes Complications 6, 39–44. Lee, A. Y., Chung, S. K. and Chung, S. S. (1995). Demonstration that polyol accumulation is responsible for

S. B A N D I T E L L I E T A L.

diabetic cataract by the use of transgenic mice expressing the aldose reductase gene in the lens. Proc. Natl. Acad. Sci. USA 92, 2780–4. Linklater, H. A., Dzialoszynski, T., McLeod, H. L., Sanford, S. E. and Trevithick, J. R. (1986). Modelling cortical cataractogenesis VIII : effects of butylated hydroxytoluene (BHT) in reducing protein leakage from lens in diabetic rats. Exp. Eye Res. 43, 305–13. Manius, G., Mahn, F. P., Venturella, V. S. and Senkowski, B. Z. (1972). GLC determination of sorbitol and mannitol in aqueous solutions. J. Pharm. Sci. 61, 1831–5. Marini, I., Bucchioni, L., Borella, P., Del Corso, A. and Mura, U. (1997). Sorbitol dehydrogenase from bovine lens : purification and properties. Arch. Biochem. Biophys. 340, 383–91. Raskin, P. (1992). Aldose reductase inhibitors—hope or hype. J. Diabetes Complications 6, 69. Robison, W. G. Jr, Nagata, M., Laver, N., Hohman, T. C. and Kinoshita, J. H. (1989). Diabetic-like retinopathy in rats prevented with an aldose reductase inhibitor. Invest. Ophthalmol. Vis. Sci. 30, 2285–92. Sarges, R. and Oates, P. J. (1993). Aldose reductase inhibitors : recent developments. Prog. Drug Res. 40, 99–161. Sato, S., Mori, K., Wyman, M. and Kador, P. F. (1998). Dosedependent prevention of sugar cataract in galactose fed dogs by the aldose reductase inhibitor M79175. Exp. Eye Res. 66, 217–22. Sorbinil Retinopathy Trial Research Group (1990). A randomized trial of Sorbinil, an aldose reductase inhibitor, in diabetic retinopathy. Arch. Ophthalmol. 108, 1234–44. Srivastava, S., Chandra, A., Bhatnagar, A., Srivastava, S. K. and Ansari, N. H. (1995). Lipid peroxidation product, 4-hydroxynonenal and its conjugate with GSH are excellent substrates of bovine lens aldose reductase. Biochem. Biophys. Res. Commun. 217, 741–6. Stevens, V. J., Rouzer, C. A., Monnier, V. M. and Cerami, A. (1978). Diabetic cataract formation : potential role of glycosylation of lens crystallins. Proc. Natl. Acad. Sci. USA 75, 2918–22. Tomlinson, D. R., Willars, G. B. and Carrington, A. L. (1992). Aldose reductase inhibitors and diabetic complications. Pharmac. Ther. 54, 151–94. Unakar, N. J., Tsui, J. Y. and Johnson, M. J. (1992). Effect of aldose reductase inhibitors on lenticular dulcitol level in galactose fed rats. J. Ocular Pharmacol. 8, 199–212. Vander Jagt, D. L., Kolb, N. S., Vander Jagt, T. J. et al. (1995). Substrate specificity of human aldose reductase : identification of 4-hydroxynonenal as endogenous substrate. Biochim. Biophys. Acta 1249, 117–26. Varma, S. D. and Kinoshita, J. H. (1974). The absence of cataracts in mice with congenital hyperglycemia. Exp. Eye Res. 19, 577–82. Wolff, S. P. and Crabbe, M. J. (1985). Low apparent AR2 activity produced by monosaccharide autoxidation. Biochem. J. 226, 625–30.